Needfreestuff, I understand and appreciate that your intention is to help, not to argue. So is mine! And in an ideal situation I'd agree with you that it makes sense to have mains and subs in-phase in the crossover region.
In audiomaze's situation, wherein it sounds like the issue may be too much energy contributed by the subwoofers too high up the spectrum, and where the minimum lowpass filter frequency setting (50 Hz) is not all that low, imo a bit of destructive interference can make a positive contribution.
Another such situation is where the mains have an intentionally exaggerated upper bass response to compensate for a lack of actual low bass response. This is sometimes the case with mini-monitors. Once again, a bit of destructive interference can cancel out that upper-bass emphasis.
A third such situation is where we're trying to squeeze a little bit more low bass out of our sub(s). If we don't have EQ available and can't reduce the lowpass filter frequency any further, sometimes we can use a bit of destructive interference to reduce the top end of the sub(s). So instead of raising the bottom end of the sub(s), we're reducing the top end. Then we can increase the level a little bit to make up for the destructive interference with the bottom end of the mains, and in effect we have raised the bottom end of the subs relative to the rest of the spectrum.
There is even a small theoretical advantage to overlapping with the mains a bit and dialing in some phase differential to smooth the response in that region: We now have more in-room sources spread around the room in the overlap region, and the net result is likely to be a bit smoother in-room response over a wider range of listening positions.
And since the ear's time-domain resolution is poor in the subwoofer region, there is unlikely to be a downside to the subs not being perfectly in-phase with the mains through the crossover region, as long as the net frequency response is good. In the bass region, speakers + room = a "minimum phase system", such that the frequency response and time-domain response track one another. When we have fixed one, we have simultaneously fixed the other.
Duke
